DERIVATIVES OF ACE~TYL HYDROXY FURANS
Jan., 1946
Decarboxylation of the dihydro acid IIIa gave the corresponding dihydro furan derivative as a COlOrkSs Oil which could not be crystallized.
Summary
89
lene-Bacetoacetate (I) by alkali to A1#*’-2’-keto3,4-dihydro-l,2-cyclopentenonaphthalene (11). Derivatives of naphtho [ 1.241furan-3-carboxylic acid also were prepared from the keto ester I by means of acid cyclization.
Conditions have been developed for the cyclization of ethyl 1-keto-tetrahydronaphtha- MADISON 6, WISCONSIN
RECEIVED
SEPTEMBER 24, 1945
I CONTRIBUTION PROM THE LABORATORY OF ORGANIC CHEMISTRY OF THE UNIVERSITY OF WISCONSIN]
Derivatives of Acetyl Hydroxy Furans and Their Conversion into Substituted Acetones1 B Y A. L. WILDS,WARRENJ. CLOSEAND JAMES A.
Several years ago the observation was made2 that 1-acetyl-2-hydroxydihydrophenanthro [1.2-b ] furan (I) wazi converted by heat into 2-phenanthreneacetone (11) in 79% yield. I n view of the unusual nature and facility of this transformation, we became interested in determining if the reaction could be extended to other acetyl hydroxyfurans. The present paper describes the results obtained with the compounds I11 and VII. The naphtho [1.2-b]furan derivative I11 was prepared fromi 1-tetralone in 73y0 yield by reactions similar to those used for the synthesis of I, namely, condensation of 2-bromo-1-tetralone with sodio-acetoacetic ester3 and cyclization to I11 by means of sodium ethoxide. The naphthofuran derivative, like I , underwent a smooth reaction upon heating a t 210-220’ and gave 2-naphthaleneacetone (IV) in 84% yield. This transformation also proceeded in good yield (74%) when 111 was heated in acetic acid solution.
I
I1 ?H
IV -
I11
a *m OH
CHI
OA>-CCEHg I
&OOCHg I /
HC1
CHaOH
V
VI
(1) This work was supported in part by a grant from the Wisconsin Alumni Research Foundation. (2) Wilds.THISJOURNAL, S4, 1421 (1942). (0) Wilds and J. A. Johnson, Jr., W.. W, 86 (1Q46).
JOHNSON, JR.
The structure of the ketone was established by synthesis from 2-naphthaleneacetic acid by conversion into the acid chloride and condensation with sodio-malonic ester. Hydrolysis and decarboxylation of the product by means of acid gave 2-naphthaleneacetone (IV),4 identical with the sample obtained from the furan derivative 111, as shown by the melting points and mixed melting points of the ketone, 2,4-dinitrophenylhydrazone and picrate. I n the conversion of the furan I11 to the ketone IV by either procedure the ketone was accompanied by a small amount (2 to 8%) of an alkalisoluble by-product which proved to be the dehydrogenated acetyl hydroxyfuran derivative V. Unlike the dihydro derivative, V was stable toward heat. Its structure follows from the analytical results and from the fact that it was converted into the methyl ester of 2-methylnaphtho[1.2-b]furan-3-carboxylicacid (VI)3by long heating with methanol containing hydrogen chloride. A similar reaction has been found to take place in the dihydro The phenanthro [4.3-b]furan derivative VI1 was prepared from 4-ketotetrahydrophenanthrene by a series of reactions similar to those used for I and 111. The conversion to 3-phenanthreneacetone (VIII) was found to go smoothly in this case, too; by heating VI1 to 200’ the ketone was formed in 81% yield, and by the acetic acid procedure, in 71% yield. The structure of the ketone VI11 was proved by synthesis from the acid chloride of 3phenanthreneacetic acid and malonic ester in a manner similar to the preparation of IV. I n this series, too, a small amount of the dehydrogenated acetylhydroxyfuran IX accompanied the ketone when the latter was prepared from the furan VII. The structure of IX was supported by its conversion into the methyl ester of the corresponding furan acid X,?when heated with methanol and hydrogen chloride. I n connection with the formation of the de(4) For this series of reactions compare Wilds and Beck, ibid.. 66, 1688 (1844). (6) For the use of this reaction in connection with the structure of I see ref. 2. (6)Borsche and Feb,Bcr., SS, 1809 (1906). (7) Wifdr and C l w , TEISJOURNAL, 6 8 , @ (1846).
A. L. WILDS,WARRENJ. CLOSEAND
90
JAMES
A. JOHNSON, JR.
VOl. 68
COCHs CHi I
CHtCOCHI
I
/COCHa
D
R
0
H
XIIIa, R = H XIIIb, R = CHaO
R
IX -
X
acetylhydroxydihydrof uran derivative into the arylacetone is supported by the additional evidence presented here. This mechanism involves rearrangement of the double bond in A away from the furan ring to form an intermediate of the type B, which can undergo aromatization by opening of the furan ring. The resulting &keto acid C would then undergo decarboxylation to the ketone D
hydrogenated furan derivatives I X and V, present evidence indicates that they do not arise as byproducts during the formation of the aromatic ketones. Instead they seem to be the result of an entirely different type of decomposition of the dihydro derivatives VI1 and 111,which takes place slowly even a t room temperature. When the dihydro derivative VI1 was heated with palladiumcharcoal catalyst in an effort to favor the formation of IX, little if any increase in yield was observed. On the other hand, the amount of I X 1 that could be isolated was increased when old CHCOCHi samples of VI1 were used. This type of decomposition was studied in more detail for the naphthofuran derivative 111. From I COZ a sample which had been allowed to stand for a D C year a t room temperature, it was ossible to isolate by crystallization about 1 5 h of the deSuch a shift of the double bond is possible for hydrogenated derivative V. Of the remaining all three of the compounds known to undergo this material approximately one-fourth no longer was transformation (I, I11 and VII). For those cases soluble in alkali. However, this neutral fraction in which it is not possible for the double bond to was not 2-naphthaleneacetone but contained 2- shift and allow the furan ring to be opened in this methylnaphtho[1.2-b]furan (X1)a corresponding manner, the transformation does not take place. to 12% of the original hydroxyfuran. No other Thus, the dehydrogenated derivatives V and IX crystalline material could be isolated from the and the examples XIIIa' and XIIIbs are reladecomposition products. The development of tively stable toward heat. It is hoped that furthe odor of acetic acid is characteristic of this slow ther work will establish more completely the decomposition. This acid could not arise from general nature of this transformation and supply the formation of either V or XI, but in the case new examples to delineate its scope and usefulof the decomposition of the phenanthrofuran ness. derivative VII, there is Bome indication that the In connection with the furan derivatives I, etc., hydroxyfuran XI1 is also produced. The full we have indicated the acetylhydroxyfuran strucnature of this slow decomposition is not yet clear, ture as the most probable. However, it is posbut it is characteristic (although a t varying sible that these compoundsmay have the alternate rates) not only of 111, VI1 and I, but also of structure of unsaturated lactones in which the derivatives of the type of XIIIa' and XIIIb,* acetyl group is in the enolic form, a formulation where the forination of a dehydrogenation prod- preferred by Borsche for the simpler compound uct (or a ketone such as IV) is not possible. XIIIa.' We hope to obtain evidence on this The mechanism suggested earlier2 to explain point by a comparison of the ultraviolet absorpthe possible course of the transformation of an tion spectra of these compounds and appropriate derivatives. (8) Wilds uad T.L. Johnlon, Tnrs JOURNAL, W, 198 (1046).
D-cH2cocHa +
DERIVATIVES OF ACETYLHYDROXY FURANS
Jan., 1946
ExperimentalQ 3-Acetgl-Z-hydrary-4,S-dihydronahtho [1.Z-b]furan
(III).-2-Bromotetralone was prepared from 20 g. of 1tetralone and 21.5 g. of bromine and converted into the acetoacetic ester derivative as described previously.' The reaction mixture was cooled, treated with a solution of 7.6 g. of sodium in 160 cc. of absolute alcohol and allowed to stand at room temperature for thirteen hours. Water was added to the gelatinous mixture, the insoluble sodium &It of the hydroxyfuran was filtered off and washed with benzene. The dried salt was dissolved in warm 60% methanol and the solution acidi6ed with dilute hydrochloric acid to precipitate the hydroxyfuran; B.3 g. of material with them. p. 133-136' was obtained, corresponding to an over-all yield of 73% (from tetralone). Recrpstallization from methanol gave light yellow blades with the m. p. 137-138.5'. The compound gave a brown solution with sulfuric acid and a green color with alcoholic ferric chloride solution. Anal. Calcd. for C I ~ ~ O :C,: 73.7; H, 6.3. Found: C, 73.6; H, 5.4. Decomposition of X I I at Room Temperature.-This decomposition occurred gradually, but was appreciable in one month. After a sample (m. p. 135.6-137') had been stored in the dark, a t room temperature (20 to 350) for one year, it had changed to a dark, red-brown mixture of liquid and semi-solid gum with a strong odor of acetic acid. Crystallization of this material (16.8 g.) from methanol gave 1.25 g. (m. p. 169-171 ') and 0.85 g. (m. p. 175-177') of a light tan solid. Upon further recrystallization from methanol the melting point of the creamcolored blades was raised to 176.5-177 This material was stable toward heat (was unchanged upon evaporative distillation at 0.2 mm.) and agreed in analysis for the dehydrogenated 3acetyl-2hydroxynaphtho [l,%b]fiiran (V). The compourid gave a greenish-yellow solution with sulfuric acid and an intense blue color with alcoholic ferric chloride solution. It was soluble in dilute sodium bicarbonate. Anal. Calcd. for CIJSIOO::C, 74.3; H, 4.5. Found: C, 74.4; H, 4.4. The original filtrate from the solid was dissolved in benzene and ether and extracted five times with 5% sodium bicarbonate solution, five times with 5% potaskium hydroxide and then washed with water. Acidification of the bicarbonate extract and treatment of the resulting oil with methanol gave an additional 0.45 g. of the dehydrogenated acetyl hydroxyfuran, m. p. and mixed m. p. 174176', bringing the total yield of this material to 15%. No additional solid could be isolated from the remaining 3.5 g. of this fraction, even after evaporative distillation (approximately one-half was non-volatile a t 250' and 0.4 mm.). Acidification of the potassium hydroxide fraction gave 7.0 g. of a dark colored oil which decomposed to a non-volatile pitch when heated. The neutral fraction contained 3.6 g. of a dark oil which gave 2.0 g. of a yellow liquid upon evaporative distillation a t 105-120' (0.2 mm.). This could not be crystallized directly, but upon treatment with picric acid in absolute alcohol gave a total of 3.7 g. of an orange picrate, m. p. 113-114'. This proved to be the picrate of 2-methylnaphtho[1.2-b]furan XI* (mixed m. p. 113-114') and corresponded to a 12% yield. A sample of the free naphthofuran was prepared from the picrate and recrystallized from methanol (using Dry Ice); m. p. 22.5-23'. The trinitrobenzene derivative, obtained as light yellow needles, had the m. p. and mixed m. p. 132.5-133.5O.' Conversion of V to Methyl 2-Methylnaphtho[1.2-b]furan3-carboxylate (VI).-Methanol (30 cc.) containing 0.3 g. of the dehydrogenated acetylhydroxynaphthofuran V was saturated with hydrogen chloride and refluxed for three days. The mixture was treated with benzene and the extract washed with water and 570 potassium hydroxide. The neutral fraction crystallized readily from methanol to give a total of 0.26 g. (82%) of the methyl ester of the furan acid VI, m. p. 61-62.5'. Recrystallization gave
'.
(9) All melting pints a r e corrected unleaa otherwise indicated.
91
colorless prisms of m. p. 63-64.5' which did not depress the m. p. of the sample prepared previously.' Similar treatment of the dihydroacetylhydroxyfurafuran derivative I11 for four hours followed by alkaline hydrolgsis gave the correspondingdihydrofuran acid' in 50% yield, m. p. and mixed m. p. 230-232'. 2-Naphthaleneacetone (IV). ( a ) From the Hydroxy Furas 111.-Four grams of the hydroxynaphthofuran 111 was melted in a 10-cc. Claisen flask and heated in an oilbath at 210-220' under slightly reduced pressure for one and one-half hours. Then the pressure was lowered slowly until the product began to distill and finally was lowered to 0.5 mm. until a higher boiling fraction began to appear. The distillate amzunted to 2.72 g. (84%) of After two r e c r y s t m a solid with the m. p. 32-35.5 tions from petroleum ether the melting point of the colorless leaflets was 37-37.5'. The material gave no color with sulfuric acid. Anal. Calcd. for CISHISO: C, 84.7; H, 6.6. Found: C,84.5; H, 6.5. Evaporative distillation (at 0.5 mm.) of the residue h theodistillation flask gave 0.1 g. of solid melting at 167200 This material appeared to contain some of the dehydrogenated acetylhydroxyfuran and gave a bluegreen color with alcoholic ferric chloride solution. When 5 g. of the hydroxydihydronaphthofuran I11 was heated under r d u x with 200 cc. of acetic acid*in a carbon dioxide atmosphere for eighteen hours, the main product agam was 2-naphthaleneacetone. The mixture was diluted with water, extracted with benzene and the latter washed with dilute alkali and water. The residue after removal of the benzene was crystallized from pztroleum From ether giving 2.48 g. of material melting a t 34.5-36 the filtrate after evaporative distillation a t 0.5 mm. and recrystallization from petroleum ether was obtained an additional 0.52 g., m. p. 33-34.5', bringing the total yield to 74%. Acidification of the alkaline wash gave 0.4 g. of solid melting a t 152-158'. The melting point was raised to 167-170' by purification through the sodium salt (using bicarbonate solution). A mixture of this material with the dehydrogenated hydroxyfuran V obtained above (m. p. 175-176') melted at 171-173.5'. The 2,4-dinitrophenylhydrazone of 2-naphthaleneacetone was obtained in nearly quantitative yield by heating the ketone (by either procedure above) with a solution of the reagent in alcohol containing an equivalent of hydrochloric acid. Recrystallition from a mixture of ethyl acetate and alcoFl gave orange hexagonal plates with the m. p. 172.5-173 Anal. Calcd. for C1~Hl~04N,: C, 62.6; H, 4.4. Found: C, 62.7: H. 4.4. -The picrate of the ketone was obtained from absolute alcohol as yellow rectangular plates melting a t 79.580'. Anal. Calcd. for C I ~ H ~ ~ O C & Q N IN, : 10.2. Found: N, 10.0. (b) From t N u ~ h t h a l e n ~ a c c t uAcd-2-Naphthaleneacetic acid (m. p. 141.6-143 ') was prepared from 2-acetylnaphthalene by the modsed Willgerodt procedure described by Schwenk and Bloch.lo The acid chloride was prepared from 7 g. of the acid, condensed with sodiomalonic ester and hydrolyzed with acid following the procedures used for 1-keto-2-methyltetrahydrophenanthrene2-acetone.' Evaporative distillation of the neutral fraction a t 0.5 mm. followed by recrystallization from petroleum ether gave a total of 3.9 g. (56%) of 2-naphthaleneacetone, m. p. 35.5-37.5"; 14% of the starting acid was recovered from the bicarbonate extract. Further -:r crystallization of the ketone raised the m. p. to 37-37.5 The 2,4-dinitrophenylhydrazone melted a t 172.5-173 ' and the picrate at 79.5-80'. The melting points of the ketone and its derivatives were undepressed when mixed with the correspondingsamples prepared in (a). 3-Acetyl-2-hydroxy-4,S-dihydrophenanthro [4.3-b]furan (VII).-A solution of 5 g. of ethyl 4-keto-1,2,3,4-tetra-
.
.
.
.
.
(10) Schwenk and Blach, Tms JOURNAL, 64, 3051 (1942); Newman, J . &#. Ckkrm., 9, 521 (1844).
92
A. L. WILDS,WARWNJ. CLOSEAND J m s A.
JOI~NSON, JR.
Vol. 68
3-Acetyl-2-hydroxyphenanthro [4.3-b]furan (IX) .-An hydrophenanthrt3ne3-acetoacetate7in 20 cc. of dry benzene was added to a solution of 1 g. of sodium in 20 cc. of ab- attempt was made to increase the amount of dehydrosolute alcohol. The red-orange mixture was allowed to genation of VI1 to I X by heating 1 g. of the material stand for twelve hours at room temperature, during which (nine months dld, m. p. 165-168") and 0.1 g. of palladiurntime only a small amount of solid separated. The mix- charcoal catalyst11 to 200-300" for one-half hour. The ture was acidified with 5% hydrochloric acid and the amount of IX isolated by extraction with sodium hydroxyellow solid filtered, giving 3.67 g. (86%) of VII, m. p. ide and sublimationoat 220-230" (0.1 mm.) was 0.08 g. Although t h i s was somewhat more 173-175" (gas). No pure material was obtained from the (8%), m. p. 220-222 benzene layer. Recrystallization of a portion of the solid than was obtained from the pure furan derivative, it was from benzene gave the analytical sample as fine, yellow not significantly larger than that resulting from heating needles with the m. p. 172.5174" (gas). A pyridine solu- an aged sample. The analytical sample was obtained as colorless needles, tion of the compound gave a green color with alcoholic ferric chloride and the solid gave a deep brown color with m. p. 222.5-223.5". The solid gave only a pale yellow concentrated sulfuric acid. Upon standing for several color with sulfuric acid but a solution in pyridine gave an months the compound became dark and underwent de- intense green color with alcoholic ferric chloride. composition. A d . Calcd. for ClaHlnOI: C, 78.3; H, 4.4. Found: Anal. Calcd. for ClsHl4Os: C, 77.7; H, 5.1. Found: C, 78.5; H, 4.3. C, 77.4; H, 5.2. From the "neutral" fraction a small amount of material 3-Phenanthreneacetone (VIII) . ( a ) From the Acefyl- believed to be 2-hydroxyphenanthro [4.3-b]furan (XII) hydroxyfuran VII.-One gram of VI1 was heated a t 180- could be isolated by heating the oil repeatedly with 5% 200" for ten minutes and then was evaporatively distilled sodium hydroxide and recrystallizing the acidic material at 0.1 mni. and the same temperature. The oily distillate several times from alcohol; yield 0.03 g. (3%), m. p. 183solidified and was recrystallized from methanol to give 184". The material gave no depression in melting pomt 0.61 g. of the ketone melting a t 70-71.5" and an additional when mixed with the sample obtained previo~sly.~ Conversion of IX to the Furan Ester X.-Forty milli0.07 g. melting a t 64-68' for a total yield of 81%. Following the oily distillate was a small amount of solid grams of I X was heated for three and one-half days with methanol saturated with hydrogen chloride. Evaporative sublimate (200-225" at 0.1 mm.) which melted a t 222.5223.5" after recrystallization from alcohol or benzene. distillation of the neutral fraction gave 18 mg. (43%) of More of this material was isolated when crude hydroxy- the furan ester X; m. p. 133-135". The recrystallzed furan was used. This is shown below to be 3-acetyl-2- product had the m. p. 135-136' alone and when mixed with a sample of the material prepared previously.' hydroxyphenanthro [4.3-b]furan (IX). Recrystallization of the 3-phenanthreneacetone from Alkaline hydrolysis gave the corresponding furan acid with the m. p. and mixed m. p. 308-309". uncor. (Pyrex tube). alcohol gave colorless leaflets with the m. p. 70.5-71.5'. Anal. Calcd. for Cl,H140: C, 87.2; H, 6.0. Found: C, 87.0; H. 6.0. summary 3-Phenanthreneacetone also was obtained in 71% yield, The earlier observation that l-acetyl-2-h~m. p. 7@72", by refluxing a solution of the hydroxyfuran (I) VI1 in acetic acid, as described above for the naphtho- droxy-10,11-dihydrophenanthro[1.2-bl furan was converted by heat into 2-phenanthrenefuran derivative. The ketone gave a 2,+dinitrophenylhydraziae deriva- acetone (11) has now been extended to 3-acetive in 99% yield; m. p. 253-254" dec. uncor. When this tyl 2 hydroxy 4,5 - dihydronaphtho[ld-blfuran derivative was recrystallized from toluene, the product gave high analytical values for carbon (found: C, 68.0, (111) and 3-acetyl-2-hydroxy-4,5-d&ydrophenanG8.2; H, 4.6, 4.8) because of retention of the solvent even thro [4.3-b]furan (VII). These derivatives gave after drying at 80" and 0.5 mm. for twelve hours. A 2-naphthaleneacetone (IV) and 3-phenanthrenesatisfactory analytical sample was obtained by recrystal- acetone (VIII) in 84 and 81% yields, respectively. lization from a large volume of n-butyl alcohol; in each Small amounts of the corresponding dehydrocase the melting point of the orange prisms was unchanged. genated furan derivatives were formed by a slow Anal. Calcd. for C2aH~8N40,:C, 66.7; H, 4.4. Found: decomposition of I11 and VI1 a t room temperaC, 66.5; H, 4.3. ture. The fact that these dehydrogenated com(b) From 3-Phenanthreneucetic Acid.-Eight-tenths gram of this acid.' was converted into 0.36 g. (45% yield) of pounds are stable toward heat provides addithe ketone by the method described earlier' for an analo- tional support for the mechanism suggested for gous transformation; 34% of the crude starting acid was recovered. The ketone had the m. p. 70.5-71.5" alone or the conversion of the acetylhydroxyfurans into when mixed with the sample prepared in (a). The 2,4- substituted acetones. dinitrophenylhydrazone (m. p. 253-254 " dec.) similarly MADISON RECEIVED SEPTEMBER 24. 1945 6, WISCONSIN showed no depression in melting point upon admixture with the derivative described above. (11) Linstead and Thomas, J . Chcm. Soc., 1130 (1940).
.
- -
-
